During 1972, 1643 recruits were discharged from RTC, SAn Diego, for the following psychiatric reasons: 35.4% for inaptitude, 4.1% for enuresis, and 60.5% for character and behavior disorders. Comparative analyses between these 1643 recruits (non-effectives) and a control sample of 1625 (effectives) were computed for the means of biographical variables, cognitive scores, odds- for effectiveness scores, a personal history questionnaire (RTS), and a cluster of 17 items from the RTS. Significant differences were obtained for all of the variables except one. Using a multiple-regression analysis on the validation sample, four variables emerged as significantly related to the criterion (R-. 509). When applied to the cross-validation sample, the variables of the RTS cluster, GCT, MECH, and expulsions or suspensions from school resulted in a cross validity of .451. The most powerful predictor was the RTS cluster, which accounted for 23% of the unique variance.